This paper describes a high-speed:parallel clustering method that is s
uitable for use as a preprocessor for motion picture compression and c
oding. In the parallel clustering process, plural pixels are sampled i
n one frame of an image. Clusters are gathered according to their char
acteristics, among them, colors, motion vectors, etc., as a way to org
anize clusters or modify cluster-parameters in parallel. This operatio
n is repeated until the frame is segmented optimally by a set of clust
ers. We propose the subframe parallel approach in order to enhance the
parallelism of the clustering method and avoid load-imbalance in para
llel tasks. The effectiveness of this approach is also presented and i
t is shown that a real-time clustering of motion picture images is ava
ilable in most cases.